Determining how best to allocate scarce funds is crucial to the action plans that governments and conservation organizations are developing to prevent species from going extinct. A Return on Investment (ROI) approach that explicitly incorporates the costs, benefits and risks of investment has been shown to improve the effectiveness and efficiency of prioritization schemes, yet has not, to date, been applied to single species. This PhD thesis examines how an ROI approach can be applied at species level, using African elephants as a case study.

In order to model the effects of investment, an important first step is being able to estimate current levels of spend accurately. I used annual spend data, collected for 192 protected areas across sub-Saharan Africa, to assess how the level of spend varies in relation to contextual factors. I found spend per unit area was significantly higher in smaller protected areas that had more range-restricted species, contained rhinos, were managed or received support from non-profit organizations, and were located in more effectively governed countries. I used the model, which explains 78% of the variation in annual spend per unit area, to predict spend across protected areas without data. From this I estimated that a total of US$367 million is spent annually on the management of protected areas in sub-Saharan Africa, with a median spend of US$185 km$\mathrm{<mrow\; data-mjx-texclass="ORD">-2</mrow>}$yr$\mathrm{<mrow\; data-mjx-texclass="ORD">-1</mrow>}$.

Secondly, identifying the level of spend needed to conserve wildlife populations, together with the enabling conditions and barriers, is vital in helping decision-makers improve the allocation of limited financial resources. I collected elephant census data for 90 protected areas (>40% African elephant population) and assessed how population trends vary in relation to level of spend and other contextual factors. I found a strong positive relationship between elephant population growth rate and level of spend (km$\mathrm{<mrow\; data-mjx-texclass="ORD">-2</mrow>}$yr$\mathrm{<mrow\; data-mjx-texclass="ORD">-1</mrow>}$) and a smaller positive effect of government effectiveness. Using the best fit model, which explains 53% of the variation in population trends, I estimated that a median spend of US$1,085 km$\mathrm{<mrow\; data-mjx-texclass="ORD">-2</mrow>}$yr$\mathrm{<mrow\; data-mjx-texclass="ORD">-1</mrow>}$ is necessary to stabilize elephant populations within my sample of 90 protected areas. I compared this to current levels of spend and found that 78% of protected areas experienced spend deficits; and overall funding must be increased by an estimated US$1.5 billion yr$\mathrm{<mrow\; data-mjx-texclass="ORD">-1</mrow>}$ if elephant numbers are to be stabilized across all sampled sites.

Finally, I developed an ROI framework for single species-based priority setting and tested it by evaluating the optimal allocation of funds across protected areas in sub-Saharan Africa under different objectives centred around maximizing (i) the number of elephants and (ii) the number of stable populations at the end of five years of differing levels of external investment (on top of current government spending). I found that the ROI framework substantially increased the number of elephants (n >18,000) or stable populations (n =28) compared to the current external investment. However, there was a considerable trade-off between the number of additional elephants and additional stable populations, particularly at lower levels of external investment. This was driven by the interplay of several factors: (i) the shape of the species response curve to investment, (ii) population sizes in each protected area, (iii) the sizes of the protected areas, (iv) the existing levels of investment, and (v) the available budget. These factors are often not considered in conventional priority-setting approaches, to the detriment of efficient resource allocation.

This thesis demonstrates how an ROI framework could be used by governments and conservation organizations that finance species conservation to evaluate the impact that increasing or decreasing their overall investments would have on species recovery. Based on the findings of this thesis, I provide several recommendations and possible solutions to improve the allocation of conservation resources, and thus support more informed and better justified investment in the conservation of species